Tube-based MPC of hierarchical distribution network for voltage regulation considering uncertainties of renewable energy

Abstract

• A novel model for low-voltage distribution areas based on genetic algorithm is developed, which simplifies the complexity of online computation and suppresses the inner voltage deviations in a safe section. • The coordination of PVs and ESs at different voltage levels is promoted to improve the voltage regulation while short of the reactive power capacity of PVs. • A tube-based MPC controller is proposed for voltage fluctuation under uncertainties introduced by power disturbance and a genetic-algorithm-based equivalent model of low-voltage areas, with which a distribution network can work in common conditions under variable operations of PVs. With the ever-growing penetration of distributed generators (DGs), voltage deviation caused by randomness and fluctuation has become increasingly prominent in distribution networks. Conventional voltage control cannot sufficiently utilize the rapid and flexible characteristics of DGs. In this paper, a tube-based model predictive control (TMPC) method is proposed to address voltage violation and fluctuation difficulties considering the uncertainties of renewable energy. Firstly, a simplified model approach based on a genetic algorithm (GA) is proposed to unify numerous DGs in low-voltage distribution areas. Then, based on this simplified model combined with the voltage sensitivity model, the TMPC controller is designed for the whole distribution network, which consists of two subsidiary controllers to suppress voltage deviation caused by uncertainties. The former is used to generate a central path for different operating points of the grid, and the latter is designed to track the path in a bound tube. The proposed method can cope with the uncertainties caused by external disturbances such as drastic power fluctuations. Finally, the numerical simulation in the case studies demonstrates the effectiveness of the proposed method in accuracy and efficiency.